A rep sets up a five-step email sequence and realizes on step 3 that prospects who clicked the pricing link in step 1 should get a completely different message than those who ignored it. In every major AI sales tool today, there is no way to branch on that behavior — the sequence is a flat list, not a tree. The complaint 'customization, especially around sequences, is lacking; if statements would be awesome' appears across products, meaning this isn't a vendor oversight, it's a category-wide gap.

The reason it persists is structural. Conditional branching requires a visual flow builder with real logic state — basically a lightweight workflow engine — which is a significant engineering investment that incumbent sales tools have deprioritized because most of their buyers evaluate on ease of setup, not sophistication of logic. The buyer (sales leadership) demos the tool, sees it send sequences, and signs the contract. The user (the rep or sales ops person) discovers the if/then limitation six weeks later when they're deep in a campaign. That buyer/user split means the feedback never reaches the purchasing decision loudly enough to force change.

What's missing is specifically: branching based on email opens, clicks, or reply sentiment; the ability to route a contact to a different step based on their response; and per-step timing that varies rather than using one global interval. The closest thing reps do today is maintain two parallel sequences and manually move contacts between them — which means contacts fall through the cracks and the 'automation' still requires babysitting.

This is a business because every new campaign a sales team builds is an opportunity to apply conditional logic, and teams that sell multiple products or to multiple personas run enough parallel sequences that the manual workaround cost is measurable in rep-hours per week.

What to build

Build a standalone visual sequence builder — with a drag-and-drop branching canvas, per-step variable timing, and trigger conditions based on email open, click, or reply — that exports executable sequences via API to common sending infrastructure or runs its own sending layer.

Where to start

Start with teams using a CRM that already fires webhooks on email events (HubSpot or Salesforce) — where click and open data is already reliably flowing — so you can build the conditional logic layer on top of proven event data without needing to instrument email tracking yourself.

The hard part

The hardest early trade-off is whether to build your own email sending infrastructure or rely on third-party APIs — building it yourself gives you full control over conditional logic execution but is a massive trust and deliverability problem to solve from day one with no reputation; relying on third-party APIs means your conditional branching is only as good as the webhooks and triggers those APIs expose.

How it makes money

Monthly subscription priced per number of active sequences or active contacts in sequences, with a free tier capped at 2 active sequences to let sales ops leads test it on a real campaign before buying.

See the evidence. The complaints behind this idea, the products they came from, and similar ideas in AI Sales Assistant.

More ideas in AI Sales Assistant